Scripta Materialia 108 (2015) 35–39

Contents lists available at ScienceDirect

Scripta Materialia

journal homepage: www.elsevier .com/locate /scr iptamat

Strengthening high-stacking-fault-energy metals via parallelogramnanotwins

http://dx.doi.org/10.1016/j.scriptamat.2015.05.0391359-6462/� 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

⇑ Corresponding authors.E-mail addresses: yfzhang@bjut.edu.cn (Y. Zhang), kjzhao@purdue.edu (K. Zhao).

Yuefei Zhang a,⇑, Jin Wang a, Haiquan Shan a, Kejie Zhao b,⇑a Institute of Microstructure and Property of Advanced Materials, Beijing University of Technology, Beijing 100124, PR Chinab School of Mechanical Engineering, Purdue University, West Lafayette, IN 47906, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 19 March 2015Revised 19 May 2015Accepted 22 May 2015Available online 15 June 2015

Keywords:NanotwinsNiTwin patternsElectrodepositionParallelogram

Twins readily form during growth or deformation in metals of low stacking-fault energy. Fabrication ofhigh-density growth nanotwins in high stacking-fault energy materials is technically challenging. Wesynthesize nanotwinned Ni foils with controlled twin patterns – pulsed electrodeposition produces con-ventional parallel nanotwins while direct current electrodeposition generates a new pattern of nanotwinsin the parallelogram form. The parallelogram nanotwins confine dislocation motion in the cage of twinboundaries and lead to a superior strengthening effect in Ni.

� 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

Nanotwinned (NT) metals exhibit unique properties of ultra-high mechanical strength, enhanced ductility, superior thermalstability, and excellent electrical conductivity [1]. The attractivephysical properties stem from the nature of R3{111} coherenttwin boundaries (CTBs) and the characteristic size at nanoscale[2,3]. CTBs with reduced spacing lead to the strengthening effectthat is analogy to the classical Hall–Petch behavior. Meanwhile,nanoscale twins enhance the deformation ductility by retainingample room to accommodate the dislocation accumulation. Inaddition, a small fraction of R3{112} incoherent twin boundaries(ITBs) often coexist with CTBs in nanotwinned materials [4,5].Propagation of ITBs contributes additional ductility of NT metalsunder a field of stress [5]. Despite the attractive mechanical prop-erties, prior studies on NT metals focused on materials of lowstacking-fault energy (SFE) or a low ratio of the unstabletwin-fault energy (cutf) to the unstable stacking-fault energy (cusf)[6]. For high-SFE metals, dislocations are the predominant carriersto mediate the plastic deformation, although twins are occasion-ally activated at extreme conditions such as at crack tips, duringhigh-pressure torsion or ball milling, and in deformed nanocrys-talline grains [7–10]. Conventional processes to fabricatehigh-density growth nanotwins in low-SFE metals include elec-trodeposition, magnetron sputtering, electron-beam evaporation,plastic deformation, and phase transformation [11–14]. Synthesisof high-density growth twins in metals of high SFE has been

extremely challenging. A few efforts were made to fabricategrowth nanotwins in Ni/Cu multilayers or by adding alloying ele-ments in the plating bath to facilitate the nucleation of twin seeds[15,16]. Successful synthesis of high-density nanotwins in puremetals of high-SFE, however, is rare, which limits the utilizationof nanotwins to strengthen a broad variety of engineering materi-als [11,17].

The primary scope of this study is to report a feasible route tosynthesize bulk Ni (SFE 125 mJ/m2) with a high density of nanos-cale twins by electrodeposition. Twin orientations can be con-trolled by varying the input current – pulsed electrodepositionfabricates Ni samples with nanotwins parallel to each other ingrains, while direct current (DC) electrodeposition leads to a regu-lar two-dimensional pattern that two sets of parallel twins consti-tute parallelograms. The average grain size and twin thickness canbe well controlled by modifying a series of electrodepositionparameters. The parallelogram nanotwins, distinct from the con-ventional parallel twins in low-SFE materials, lead to ultrahighmechanical strength and tensile ductility. The enhanced mechani-cal properties of parallelogram NT samples are attributed to theconfinement effect for dislocation motion thus a larger capacityof dislocation storage and accumulation.

Table S1 summarizes the experimental conditions. Each param-eter is varied separately in a given range to test the individualeffect on the microstructure of as-deposited samples. The quantita-tive effects of the bath temperature, current density, Ni2+ concen-tration, and ratio of current on-time and off-time (Ton/Toff) on theaverage grain size, twin thickness, and twin length are illustrated

36 Y. Zhang et al. / Scripta Materialia 108 (2015) 35–39

in the Tables S2–S5. Overall, the increase of the above experimen-tal controls results in a decrease of the average grain size and anincrease of the twin density (Table S6). The explicit mechanismof nanotwinned Ni fabricated by the simple conditions of elec-trodeposition remains unclear at this writing due to the complexnature of electrochemical processes. Nevertheless, we aspire todisclose the experimental details that repeatedly produce theintriguing nanotwin patterns as summarized in the Tables S1–S6.

The as-deposited Ni foils have in-plane dimensions of 30 mm by30 mm and a thickness of 30 lm with a uniform columnar-grainedmicrostructure and high purity (Figs. S1 and S2). Here we will focuson the characterization of the parallelogram nanotwins.Information on parallel twins is included in the Supplementarymaterials. Fig. 1a shows a transmission electron microscopy(TEM) image on the morphology of parallelogram twins embeddedin submicron grains. The grain sizes are between 400 nm and800 nm. Each grain is divided into parallelograms by a high densityof growth twins. Measurements of the twin thickness along the[110] orientation show a wide distribution ranging from severalnanometers to about 120 nm. High-resolution TEM (HRTEM)observations illustrate rich microstructural features.Multiple-fold twin boundaries exist at the junctions of parallelo-gram twins to fulfill the geometric compatibility (Fig. 2c) [18–20]. A mismatch angle presents at the junction of multiple-foldtwin boundaries. In fivefold twins, for instance, the angle betweentwo neighboring {111} planes is 1.47� smaller than that requiredto perfectly cover one fifth of a circle. Such a mismatch gap isaccommodated by stretching and broadening individual twinboundaries, which may act as a source of stress concentration

Fig. 1. (a) The bright-field TEM image shows high-density parallelogram nanotwins in ashows roughly equiaxed submicron grains with random orientations. (b) Statistical distrexist at the junctions of parallelogram twins to fulfill the geometric compatibility. TheHRTEM image (inset). (d) CTBs and ITBs coexist. The percentage of CTBs and ITBs are 88

and dislocation nucleation. Multiple-fold twin boundaries alsoconnect with grain boundaries (GBs) and form a complex twin net-work structure. CTBs and ITBs also coexist (Fig. 2d). The densities ofCTBs and ITBs in parallelogram nanotwins are 4.5 � 1014 m�2 and5.8 � 1013 m�2, respectively. The microstructural features of paral-lel nanotwins are shown in Fig. S3. It is noteworthy that the densi-ties of CTBs and ITBs in parallelogram NT samples are markedlylarger than their counterparts in parallel nanotwins�2.9 � 1014 m�2 for CTBs and 1.9 � 1013 m�2 for ITBs.

Thermal stability of the growth nanotwins in as-deposited Nifoils is characterized by isothermal annealing. Detailed experimen-tal procedure is shown in the Supplementary materials. In situ TEMobservations of the microstructural evolution of parallelogramtwins annealed at 100 �C, 200 �C, 300 �C, 400 �C, 500 �C, and600 �C are shown in Fig. 2. Detwinning is observed in the samplesannealed above 400 �C.

Tensile tests of NT foils are performed at a nominal strain rate of10�4 s�1 at ambient temperature. Fig. 3 shows the high strength ofparallelogram NT samples in comparison with that of parallel NTsamples, a coarse-grained polycrystalline Ni sample, and ananocrystalline Ni sample. The microstructures of nanocrystallineand coarse-grained Ni and grain size distributions are shown inFig. S4. The parallelogram NT Ni samples also show considerabletensile ductility, with an elongation-to-failure strain of 6.5%, whichis much higher than the nanocrystalline specimen. A slight strainhardening appears in the plot of engineering stress versus engi-neering strain during the major stage of plastic deformation.Converting the plot to true stress versus true strain, the work hard-ening would become more prominent, indicative of dislocation

Ni sample synthesized by DC electrodeposition. Electron diffraction pattern (inset)ibutions of the grain size and the twin thickness. (c) Multiple-fold twin boundariesfivefold TBs are broadened to accommodate the mismatch angle as shown in the% and 12% respectively.

Fig. 2. Thermal stability of the growth nanotwins in as-deposited Ni foils. In situ TEM observations of the microstructure of parallelogram twins isothermally annealed at100 �C, 200 �C, 300 �C, 400 �C, 500 �C, and 600 �C respectively. Detwinning is observed in the circled areas annealed above 400 �C.

Fig. 3. Stress–strain curves for the as-deposited Ni samples with parallel andparallelogram nanotwins (with typical twin thickness of 30 nm and 50 nmrespectively, the average of grain sizes are both in between 300 and 500 nm) incomparison with that for a coarse-grained polycrystalline Ni sample (with anaverage grain size of ca. 400 nm) and a nanocrystalline Ni sample (with an averagegrain size of ca. 20 nm). Each type of nanotwinned Ni is tested with two samples.The inset sketches the geometry of the samples.

Y. Zhang et al. / Scripta Materialia 108 (2015) 35–39 37

accumulation before failure [21]. Necking and shear cracking occurupon tensile fracture. The fracture surfaces (Fig. S5) demonstratedimpled rupture and microvoid coalescence. The dimple sizes arebetween a few hundred nanometers and several micrometers withan average value of a few times of the average grain size. It is note-worthy that by varying the grain size and twin thickness it mayinduce size-dependent deformation behaviors and failure mecha-nisms. Such size effects would be intriguing for further systematicstudies [22]. The difference in the elastic modulus between paralleltwinned Ni and those with parallelogram nanotwins is primarilydue to the different current densities in the fabrication of thetwo sets of samples. The tested samples with parallelogram twins

were fabricated at a current density of 15 A/dm2, while a currentdensity of 36 A/dm2 was provided for the parallel twin samples.The high current density generally leads to a less dense structurewith lower elastic modulus [23–25].

The strengthening mechanism by means of nanotwins in metalsis generally attributed to the interactions between glide disloca-tions and twin boundaries. CTBs act as a strong barrier for disloca-tion motion leading to a strengthening behavior analogous to thatof conventional GBs. HRTEM images in Fig. S6 show thathigh-density of dislocations pile-up at CTBs and transmit acrossthe twin boundaries. The interactions between dislocations withCTBs generate a large number of stacking faults. The segregationof dislocations distorts the twin boundaries. The reactions betweendislocations in face-centered cubic metals and twin boundariesgenerate glissile Shockley partial dislocations (SPDs) and sessileFrank dislocations (FDs). The glissile partial dislocations may prop-agate across the twin boundaries, while the sessile dislocations areleft at the boundaries and result in a loss of coherency of CTBs [26].Nevertheless, other dislocation reactions cannot be excluded andmay also be operative in the strengthening mechanism [27,28].Dislocations nucleated at various sites of structural defects, pri-marily at GBs and intersections with twin boundaries, slip alongthe twin planes. The successive motion of glissile dislocations leadsto the migration of CTBs that mediates the plastic flow of NT Ni,Fig. S6b [29]. In addition, ITBs are another important carrier oftwining plasticity. ITBs are highly mobile, Fig. S6c, which requiresa much smaller resolved shear stress than that required to emittwinning partial dislocations from GBs with perfect CTBs [4].Detwinning is accompanied with the migration of ITBs via the col-lective slip of multiple twin dislocations, Fig. S6d [30]. The defor-mation mechanisms of NT Ni observed in our experiments sharemuch similarity with that of NT Cu which has been intensivelystudied. This observation might be roughly understood that bothNi and Cu have low ratios of stacking fault energy (csf) to cusf

and cusf/cutf, such that partial dislocations dominantly mediate

Fig. 4. Dislocation activities in parallel and parallelogram nanotwinned Ni after applying a tensile strain of �5%. (a and c) For parallel nanotwins, dislocations are primarilylocated at the twin boundaries and few are seen inside lamellar layers. (b and d) For parallelogram nanotwins, dislocations are confined in the cage of twin boundaries and areobserved in both the vicinity and interior of twin boundaries.

38 Y. Zhang et al. / Scripta Materialia 108 (2015) 35–39

plasticity in both materials. Detailed atomistic pictures in high-SFEmetals warrant further careful studies [12].

Despite the common deformation mechanisms in the paralleland parallelogram NT Ni, the NT patterns interplay with disloca-tions in different manners. For parallel nanotwins, dislocationsare confined only by the thickness of twin lamellae while a suffi-cient free path is maintained for dislocation slip along the lengthorientation, Fig. 4a. In contrast, dislocations are confined in thecage of parallelogram twin boundaries. Such confined spacesmay be effectively regarded as sub-equiaxed-grains, Fig. 4b. Thereduced size results in more accumulative dislocations at thetwin boundaries and thereby a better strengthening effect of par-allelogram nanotwins. It is confirmed by our observations thatdislocations are located in both the vicinity and interior of twinboundaries for the parallelogram twins, while dislocations in par-allel twins primarily pile-up near the twin boundaries and feware seen inside the twin lamellae, Fig. 4c and d. Furthermore,the densities of CTBs and ITBs in the parallelogram form areinherently larger than that in the parallel twin samples. It is con-ceivable that the dislocation density in parallelogram nanotwinsis much larger than the counterparts of parallel twins upon defor-mation. We may estimate the material strength following theTaylor-type relationship, rflow �

ffiffiffiffiqp , where q is the dislocationdensity. A larger capacity for dislocation storage and accumula-tion renders higher mechanical strength and extended ductilityin parallelogram nanotwins. Beyond the Taylor-type strengthen-ing argument, it is also plausible that dislocations from multipleslip systems intersect at twin boundaries in the parallelogramnanotwins that abundant Lomer–Cottrell (L–C) locks may formto enhance the work hardening capability. The L–C locks wererecently observed within grains and at twin boundaries innanocrystalline Ni [31].

In summary, we synthesized nanotwinned Ni foils with con-trolled spacing and pattern through conventional electrodeposi-tion. The parallelogram nanotwins are distinct from the paralleltwins that are commonly observed in low-SFE metals. Such a pat-tern confines dislocation motion in the interior space of twinboundaries, and may inspire a new strategy of designinghigh-strength metals via patterned nanotwins.

Acknowledgments

Y.Z. acknowledges the National Natural Science funded project(11374027) and Special Projects for Development of NationalMajor Scientific Instruments and Equipment (2012YQ03007508).The research project is supported by the start-up funds atPurdue University. K.Z. acknowledges helpful discussions withDr. Xiangdong Ding, Dr. Shaoxing Qu, and Dr. Jian Wang. K.Z. isgrateful for the support of Haythornthwaite Foundation InitiationGrant from American Society of Mechanical Engineering.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.scriptamat.2015.05.039.

References

[1] Y. Lu, X. Shen, L. Chen, K. Qian Lu, Science 304 (2004) 422–426.[2] K. Lu, L. Lu, S. Suresh, Science 324 (2009) 349–352.[3] L. Lu, X. Chen, X. Huang, K. Lu, Science 323 (2009) 607–610.[4] Y.M. Wang, F. Sansoz, T. LaGrange, R.T. Ott, J. Marian, T.W. Barbee Jr., A.V.

Hamza, Nat. Mater. 12 (2013) 697–702.

Y. Zhang et al. / Scripta Materialia 108 (2015) 35–39 39

[5] J. Wang, N. Li, O. Anderoglu, X. Zhang, A. Misra, J. Huang, J. Hirth, Acta Mater. 58(2010) 2262–2270.

[6] H. Van Swygenhoven, P. Derlet, A. Frøseth, Nat. Mater. 3 (2004) 399–403.[7] H. Zhou, S. Qu, W. Yang, Model. Simul. Mater. Sci. Eng. 18 (2010) 065002.[8] J. Huang, Y. Wu, H. Ye, Acta Mater. 44 (1996) 1211–1221.[9] X. Wu, Y. Zhu, M. Chen, E. Ma, Scripta Mater. 54 (2006) 1685–1690.

[10] X. Liao, F. Zhou, E. Lavernia, D. He, Y. Zhu, Appl. Phys. Lett. 83 (2003) 5062–5064.

[11] H. Idrissi, B. Wang, M.S. Colla, J.P. Raskin, D. Schryvers, T. Pardoen, Adv. Mater.23 (2011) 2119–2122.

[12] D. Bufford, Y. Liu, J. Wang, H. Wang, X. Zhang, Nat. Commun. 5 (2014) 4864.[13] D. Bufford, H. Wang, X. Zhang, Acta Mater. 59 (2011) 93–101.[14] V. Yamakov, D. Wolf, S. Phillpot, H. Gleiter, Acta Mater. 50 (2002) 5005–5020.[15] Y. Liu, D. Bufford, S. Rios, H. Wang, J. Chen, J. Zhang, X. Zhang, J. Appl. Phys. 111

(2012) 073526.[16] G. Lucadamo, D. Medlin, N. Yang, J. Kelly, A. Talin, Philos. Mag. 85 (2005) 2549–

2560.[17] D. Bufford, Y. Liu, Y. Zhu, Z. Bi, Q. Jia, H. Wang, X. Zhang, Mater. Res. Lett. 1

(2013) 51–60.

[18] X. Liao, J. Huang, Y. Zhu, F. Zhou, E. Lavernia, Philos. Mag. 83 (2003) 3065–3075.[19] H. Chen, Y. Gao, H. Zhang, L. Liu, H. Yu, H. Tian, S. Xie, J. Li, J. Phys. Chem. B 108

(2004) 12038–12043.[20] Y. Zhu, X. Liao, R. Valiev, Appl. Phys. Lett. 86 (2005) 103112.[21] Y. Shen, L. Lu, Q. Lu, Z. Jin, K. Lu, Scripta Mater. 52 (2005) 989–994.[22] Z. Wu, Y. Zhang, M. Jhon, D. Srolovitz, Acta Mater. 61 (2013) 5807–5820.[23] J. Luo, M. Pritschow, A. Flewitt, S. Spearing, N. Fleck, W. Milne, J. Electrochem.

Soc. 153 (2006) D155–D161.[24] T. Fritz, M. Griepentrog, W. Mokwa, U. Schnakenberg, Electrochim. Acta 48

(2003) 3029–3035.[25] Y. Woo, S.-H. Kim, J. Mech. Sci. Technol. 25 (2011) 1017–1022.[26] B. Wang, H. Idrissi, M. Galceran, M.-S. Colla, S. Turner, S. Hui, J.-P. Raskin, T.

Pardoen, S. Godet, D. Schryvers, Int. J. Plast. 37 (2012) 140–156.[27] Z. Wu, Y. Zhang, D. Srolovitz, Acta Mater. 57 (2009) 4508–4518.[28] D. Jang, X. Li, H. Gao, J.R. Greer, Nat. Nanotechnol. 7 (2012) 594–601.[29] Z. You, X. Li, L. Gui, Q. Lu, T. Zhu, H. Gao, L. Lu, Acta Mater. 61 (2013) 217–227.[30] J. Wang, A. Misra, J. Hirth, Phys. Rev. B 83 (2011) 064106.[31] J.H. Lee, T.B. Holland, A.K. Mukherjee, X. Zhang, H. Wang, Sci. Rep. 3 (2013)

1061.